[time-nuts] The Demise of LORAN (was Re: Reference oscillator accuracy)
Magnus Danielson
magnus at rubidium.dyndns.org
Mon Nov 16 20:55:51 UTC 2009
Lux, Jim (337C) wrote:
> A *much* more effective and cheap strategy is a repeater jammer.. Receive
> the signal and retransmit it: two antennas and an amplifier. The victim sees
> the delayed retransmitted signal at a higher level than the direct one. It's
> sort of like creating fake multipath interference. No need for PN
> generators, oscillators, etc.
The delay attack is known (it's actually more of a spoofing attack), but
most jamming uses just simple CW or noise signals. A problem with the
delay attack is that you have a risc of creating a feedback loop which
creates an unstable oscillation, which may or may not be what you would
like. Among other things, directional finding on an oscillation is much
easier than the delayed signal. However, a good hint is to use a
directional L1 antenna and point it to the signal and just hook a
receiver up and the position you have is that of the attackers antenna.
Again that reveals the position. I guess this is why the attack isn't
particularly used.
> Granted, a smart receiver that *understands* the relationship between SV to
> user geometry and doppler can beat it (because the carrier phase and PN
> phase of the repeated signal won't be consistent with the geometry), but the
> run of the mill PN tracking loop probably won't.
Actually, even the simpler receivers has some resistance to it, the
higher amplitude is the main problem.
> Most inexpensive receivers use a single bit sampler, so a suitable CW tone
> could also probably jam it effectively, but might require some knowledge of
> the victim receiver to pick an appropriate frequency (i.e. You'd need to
> know the sampling rate.)
The CW attack is well known and has been analyzed fairly deeply. You
don't need the sampling rate to make it efficient. The one-bit receiver
is dead in the sea compared to even 1.5 bit receivers with suitable AGC
loop detection.
Cheers,
Magnus
More information about the Time-nuts_lists.febo.com
mailing list